On Tue, Oct 01, 2019 at 06:30:42AM -0700, Guenter Roeck wrote:
> > @@ -4082,10 +4074,9 @@ static int tcpm_pr_set(const struct typec_capability *cap,
> > return ret;
> > }
> > -static int tcpm_vconn_set(const struct typec_capability *cap,
> > - enum typec_role role)
> > +static int tcpm_vconn_set(struct typec_port *p, bool source)
> > {
> > - struct tcpm_port *port = typec_cap_to_tcpm(cap);
> > + struct tcpm_port *port = typec_get_drvdata(p);
> > int ret;
> > mutex_lock(&port->swap_lock);
> > @@ -4096,7 +4087,7 @@ static int tcpm_vconn_set(const struct typec_capability *cap,
> > goto port_unlock;
> > }
> > - if (role == port->vconn_role) {
> > + if (source == port->vconn_role) {
>
> source is boolean, vconn_role is enum typec_role.
> The original typec code took advantage of typec_role == TYPEC_SINK matching false,
> and typec_role == TYPEC_SOURCE matching true, but I don't think it is a good
> idea to carry that down to low level drivers. This will confuse everyone who wants
> to contribute a driver in the future.
OK, I'll keep the parameter as emum typec_role.
thanks,
--
heikki
